The Effectiveness of AI in Player Experience Personalization

Abstract

The effectiveness and potential of Artificial Intelligence (AI) techniques in automatically modulating various game attributes for providing dynamic gaming experiences while ensuring playability is demonstrated and defined. Specifically, the focus is on personalization of player experience for enhancing user engagement and satisfaction. The results indicate that AI-driven personalization increases the positive sentiment in game reviews and is an effective tool to increase the engagement of both well and poorly received games. The use of AI in orchestrating a wide range of gaming experiences across different games and platforms is coined as ‘Experience Driven Playing Companion,’ and the methodology and tools developed under this paradigm are introduced. The approach is centered around the use of various approaches to model player experience, both interactively and non-interactively, using psychophysiological game events as well as visual aesthetics features and crowd-sourced game review descriptors. Several AI techniques are described for driving level difficulty adaptation, audio soundtrack generation and controlling screen brightness adaptively in various game environments. Furthermore, an overview is given on the underlying software and hardware architecture, as well as on the released gaming companion platform that allows non-expert game developers to automatically create a wide range of working playing companions by taking advantage of various black-box AI techniques. Finally, the ethics and potential pitfalls of AI-driven player experience orchestration are discussed, highlighting directions for ensuring an ethical stance in future deployment of similar systems.

Keywords AI, player experience, personalization, user engagement, game reviews, experience-driven companion, adaptive gameplay, ethical considerations